matrixStats-package: Package matrixStats

Description

This packages provides methods operating on rows and columns of matrices, e.g. col/rowMedians() and col/rowRanks(). There are also some vector-based methods, e.g. anyMissing(), weightedMedians(), and binMeans(). The objective is to have all methods being optimized for speed and memory consumption.. This package went public (on CRAN) in June 2009. It is currently in a beta version where new methods are added.

Arguments

Installation

To install this package, please do: install.packages("matrixStats")

Dependancies and other requirements

See DESCRIPTION file for now.

To get started

Counts and logicals:

rowCounts()
rowAlls()
rowAnys()

Sums and products:

rowSums()
rowProds()
rowLogSumExps(), logSumExp()

Estimates of the mean:

rowMeans()
rowMedians()
rowWeightedMeans()
rowWeightedMedians(), weightedMedian()

Estimates of the standard deviation, variance and more:

rowSds()
rowMads()
rowVars()
rowIQRs()
rowQuantiles()
rowOrderStats()
rowRanks()
rowRanges()
rowMins()
rowMaxs()
madDiff(), sdDiff(), varDiff()
weightedMad()

Binning:

binCounts()
binMeans()

Miscellaneous:

rowDiffs()
anyMissing()

How to contribute

This is an open-source project which embraces collaborations. If you have improvements on code and/or documentation, or new function, please consider contributing them to this package.

For developers

It is currently not decided whether the methods should be S4 or S3 methods. This is the reason why some methods are based on S4 and some on S3. The ones using S3 rely on the R.methodsS3 package to define the methods. There are also dependancies on other packages. The plan is to remove all such dependancies as soon as the API settles, but until then, we keep the dependancies for conveniency and in order to avoid reduncancy of available implementations of identical methods.

How to cite this package

Henrik Bengtsson (2013). matrixStats: Methods that apply to rows and columns of a matrix. R package version 0.8.3.